Instructions to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit

Run Hermes

hermes

OpenClaw new

How to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

MLX LM

How to use caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Qwen3.6-27B-Opus-Distill-OptiQ-4bit

OptiQ mixed-precision MLX quant (~3.97 bpw, data-driven KL-sensitivity 4/8-bit) of TeichAI/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2 (Apache-2.0) — a dense qwen3_5 (hybrid linear-attention) distilled on Claude Opus reasoning traces. Quantized with mlx_optiq.

Profile (64GB Apple-Silicon, 256K-agentic-coding study):

Clears 256K context: mx.get_peak_memory = 43.3 GB @ 256K (4-bit KV), retrieval 1.00.
Strong single-shot: HumanEval+ 100%, AIME 100%, math500 ~81%, LCB 80%, BFCL 0.94; aider-polyglot ~75%.
Dense-27B → slower decode than a sparse MoE at long context (~9 tok/s @ 256K).

Recommended sampling (op-temp from testing): temperature 0.3, top_p 0.95, top_k 20, min_p 0, presence_penalty 0, thinking enabled (thinking_budget 81920).

MLX/OptiQ quant + evaluation by @caslca. Base model + distillation by TeichAI (Apache-2.0).

Downloads last month: 616

Safetensors

Model size

27B params

Tensor type

BF16

U32

MLX

Hardware compatibility

4-bit

Model tree for caslca/Qwen3.6-27B-Opus-Distill-OptiQ-4bit

Base model

Qwen/Qwen3.6-27B

Finetuned

TeichAI/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2

Quantized

(12)

this model